Been having issues since 2.0 version (BUGS?)

I’ve been having a few issues since the 2.0 update

So I’ve been an n8n cloud user for over a year now, and I was always super happy about it, never had any real problem. However since the 2.0 I’ve faced some issues…

  • Sometimes I need to retry a whole execution, they way I always did it was to copy the execution to editor (pinning the trigger node) and just click the Execute Workflow button… It worked like a charm, now every time it fetches data from a completely different execution… So I pin execution 5 and it uses data from execution 1… When I open the pinned trigger the informations is right as soon as the node after runs it already has the wrong information… See images:

  • The publish workflow feature (and actually the auto save aswell) instead of the old save button I find it extremely risky… Like before I could make a quick test on a production workflow and I knew I wouldn’t mess anything if I just didn’t save… Now I find my self making a quick test (deactivating some nodes for testing) and then for a whole day my client gets duplicated leads because the workflow auto saved and kept the Remove Duplicates node deactivated…

Are this intended behaviors or bugs?

Can I somehow turn off the auto save feature?

Information on your n8n setup

  • n8n version: 2.17.5
  • Database: Default
  • n8n EXECUTIONS_PROCESS setting: Main
  • Running n8n via: n8n Cloud
  • Operating system: MacOS

Hi @Joao_Marques, welcome!
Auto-save cannot be turned off, as it is a part of the system now, and for that copying issue, I believe it is very clearly stated here:

That n8n copies the execution data into the current flow and then pins it in the first node, and what you are describing as that pinned data is not being used it is very unusual, and I would recommend giving your instance a restart from the admin dashboard and then letting me know if the issue persists.

Sad auto save cannot be turned off :frowning:

About the other issue… I can’t restart it right now because we actually can’t have downtime during work hours. But I’ve restarted my instance, I’ve updated to last stable versions, multiple times and this issue persists… Literally since 2.0 on all workflows.

@Joao_Marques, that is very strange. If I clearly understood the issue, your pinned data does not get used, and instead your old data, which was previously there, gets used. I think it is worth reporting if that is really what happens. Also if possible, provide a little more context with images and sample flow, i think then you can report this

@Anshul_Namdev first of all thanks for the quick answers!

Exactly that! And I mean the data it uses it’s actually completely random, not form last execution or anything, just random from any execution except from the pinned one.

Need more images than what I provided above?

That’s strange. Have you tried this in a different flow? or file? Does it shows the same behaviour? If it does, I really recommend posting this to

i believe this only happens when you copy the execution flow, not in all runs, like manual execution run and then pinning that data

@Anshul_Namdev Yes happens in other flows, with different triggers.

I cannot post to that category btw.

And correct this only happens when I try to manually re-run an execution. Normal executions work perfectly.

@Joao_Marques since this is a bug on n8n Cloud, consider reaching out to them by mail: